论文部分内容阅读
为了解决复杂体系分析、仪器分析大数据挖掘,以及跨仪器平台数据整合等分析化学各分支领域中的复杂应用问题,设计开发了ChemPattern一站式化学计量学分析软件系统。通过标准数据接口技术实现了可用于色谱、质谱、光谱和核磁共振波谱所获得的各类高维、高分辨海量数据的信号处理分系统,包括LC-HRMS解卷积、分子特征匹配、DART-MS指纹图谱、色谱峰保留时间校正及共有模式建模等功能。并基于高性能并行计算和大规模数据可视化技术,提供对多元统计、校正模型、模式识别及人工智能等众多化学计量学应用的支持。最后,对分析系统的可靠性进行了验证。该系统已成功用于中药指纹图谱、食品安全快速筛查及代谢组学等分析领域并取得了较传统定性、定量方法更为全面和灵敏的分析结果,具有较好的应用前景。
In order to solve the complicated application problems in complex system analysis, big data mining in instrumental analysis and analytic chemistry in cross-instrument platform data integration, ChemPattern’s one-stop chemometric analysis software system is designed and developed. Signal processing subsystems for various high-dimensional, high-resolution mass data obtained from chromatography, mass spectrometry, spectroscopy and nuclear magnetic resonance spectroscopy are achieved through standard data interface technologies including LC-HRMS deconvolution, molecular feature matching, DART- MS fingerprinting, peak retention time correction and common mode modeling and other functions. Based on high-performance parallel computing and large-scale data visualization technology, it supports many chemometric applications such as multivariate statistics, calibration models, pattern recognition, and artificial intelligence. Finally, the reliability of the analysis system is verified. The system has been successfully used in traditional Chinese medicine fingerprinting, rapid screening of food safety and metabolomics and other areas of analysis and made more than the traditional qualitative and quantitative methods more comprehensive and sensitive analysis results, has a good prospect.